Also the title is inaccurate as different thorium reactors have been designed, built and operated. See MSRE and THTR-300 for example.
LFTR designs are fantastic, amazing, and solve a lot of problems. There is a lot of materials science to be done to make them viable, though.
Also, no, not any metal gets brittle; inconel and hastelloy handle radiation quite well for decades at a time, as does good old fashioned nickel. Beyond that, most reactor pressure vessels are a layer of metal then a layer of something that's really good at radiation, then the real pressure vessel, so that the interior layer's brittleness isn't very important.
We built and ran LFTRs commercially in the 1950s in New York State and Pennsylvania, before computers became a commercially realistic thing. They provided our grandparents no significant technical challenge.
The actual big problem with LFTR is primarily regulatory. The design hasn't been vetted to modern safety standards, which costs hundreds of millions of dollars, and the entities who are nuclear-aware and have that kind of money to throw around tend to be the existing nuclear companies, who can't switch to any other technology because they're deep in the Gilette Razor model, and anything that took out their existing fuel contracts would immediately bankrupt them.
There is /zero/ materials science needed to make a LFTR. I don't know where you got that idea. They're substantially easier than what we make today. The average auto body shop can pull it off.
Nope, pretty sure I'm not. I'm well aware that LFTRs run at low (even sub-atmosphere) pressures.
>We built and ran LFTRs commercially in the 1950s in New York State and Pennsylvania
We never ran LFTRs commercially. If you know otherwise, please cite. I only know of two experimental reactors at the Oak Ridge facility: the Aircraft Reactor Experiment and the Molten-Salt Reactor Experiment.
> I don't know where you got that idea.
I got that idea from some pretty simple facts about nickel alloys (like Hastelloy N) under neutron bombardment.
When you bombard nickel with neutrons, you produce helium. When the helium builds up irregularly, the alloy becomes brittle. You can dope Hastelloy N with titanium or niobium to even out the distribution of helium deposits (this is what ORNL did) but that brings the maximum temperature down to 650C.
As well, tellurium (one of the fission products of a LFTR reactor) corrodes the grain boundaries of Hastelloy N. You can reduce this effect by doping it with niobium and keeping the UF4/UF3 ratio to less than 60.
You have to trade off lower temperatures with whether or not you want to deal with beryllium toxicity. You can replace BeF2 with a eutectic lithium fluoride/thorium fluoride composition, but that requires an increased temperature of the reactor salts. There are other problems with using beryllium, though - it produces lithium-6, which is a strong neutron poison.
You also have to filter out noble element deposits, because they don't form fluorides.
There are also serious design challenges with modifying current turbines to work with supercritical CO2 or helium. You can use supercritical steam instead, but it isn't nearly as efficient.
You also have to worry about tritium diffusion. It's small enough that it leaks through the heat exchangers.
There are issues with the rapid expansion/contraction the graphite moderator, but some are working on graphite pebble designs.
Once you throw the corrosive salts, strange reaction byproducts, and neutron bombardment into the mix, I highly doubt that 'the average body shop' could pull off the fabrication of a LFTR style molten salt reactor that could run safely for longer than a week.
Everything at Oak Ridge did _not_ provide our grandparents with insignificant technical challenges. Not only for the points (in my other response) about neutron bombardment making the Hastelloy N brittle, but also:
- The cooling circuit glowed red hot. This indicates that it was close to its thermal creep zone; we didn't know much about thermal creep back then. [0]
- The MSRE reactor spent a good portion of its life down for maintenance. [0]
- The tritium release is not a small problem: of an estimated 2.0TBq of tritium produced, 6-10% diffused from the fuel system into the containment cell atmosphere, and another 6-10% was released from the heat removal system. And these were lower than expected. You really don't want your LFTR to be leaking tritium everywhere. [1]
- The decontamination of the experiment was really, really hazardous. Not only were they dealing with nuclear waste; they were dealing with nuclear waste and fluorine gas. You really don't want to be anywhere near elemental fluorine gas. This was all because ORNL didn't defuel and store the salts correctly; it's mostly been rectified in modern designs[1]. But it goes to show how horrifyingly dangerous the components are. As bad as a nuclear spill would be, you have to remember that these aren't your run of the mill molten salts. They're highly toxic salts spewing out gamma, beta, and alpha radiation. I'm trying to imagine what it would be like if Chernobyl released these kinds of chemicals in its radiation plume. I don't think I can imagine something that horrifying. Toxic rain over Germany doped with uranium hexafluoride and plutonium? I'd rather have coal power. And I hate coal power.
Of course, all of this is me being extraordinarily cautious. I was really, really excited about LFTR design when I first heard about it. But the more I learn, the more I'm convinced that we really don't have the materials to safely operate one of these things on a commercial scale.
0. http://daryanenergyblog.wordpress.com/ca/part-8-msr-lftr/ 1. http://en.wikipedia.org/wiki/Molten-Salt_Reactor_Experiment#... (throughout)
http://www.carbontax.org/blogarchives/2013/11/21/why-officia...
http://www.forbes.com/sites/energysource/2014/02/20/why-the-...
I've read about this before. Companies start building plants based on existing regulations. Along the way, the NRC changes the regulations, requiring tear-down and rebuilding. Meanwhile interest on the loan keeps building up. Add further delays due to political resistance and it's no wonder costs escalate.
It's not just the economics of big engineering causing the problem here. I think matters have improved somewhat in the U.S., but it's still going to be interesting to see how AP-1000 costs in the U.S. compare to those in China.
Incidentally, there are some arguments that liquid thorium reactors, and some other GenIV designs, could have significantly lower capital costs than conventional reactors. A big reason for that is that the basic physics of fuel and coolant provides substantial passive safety, rather than relying on lots of redundant active systems.
Aside from that, nuclear reactors don't necessarily have to be gigawatt-size. Even in the U.S., smaller reactors are starting to make some regulatory headway.
But even conventional 1GW reactors don't have to take decades to build. China's first AP-1000s at Sanmen are being finished up this year and next for a construction time of five years.
http://en.wikipedia.org/wiki/Sanmen_Nuclear_Power_Plant
http://www.world-nuclear-news.org/NN-First-Haiyang-AP1000-ta...
http://en.wikipedia.org/wiki/Thorium_fuel_cycle#List_of_thor...
It makes me sad that it's not happening in the US. If some folks in the government and business don't get their heads out of their asses a growth industry is going to bypass us entirely.
As India is already a nuclear weapons power, this has no immediate proliferation implications ... but moving a nation of a billion-plus people onto an energy cycle that produces weaponizable material as a by-product might be considered unwise by some. Cf. concerns in the 1970s and 1980s about the implications of running a "plutonium cycle" fast breeder energy ecosystem.
I'm not saying it's a slam-dunk win for sure.
But I have noticed a lot of general apathy and aversion to violence in the developed world largely because people are just too busy living their lives; they have a lot to lose.
> the real question is how they plan to reduce the U-232 contamination level enough to make weapons-safe U-233
I'm not convinced that anyone has this goal in mind. Maybe they just want to provide power to their countrymen and continue to lift India out of poverty. There might be nothing nefarious about this, unless you consider poor people getting less poor to be a problem.
I mean, the local coal power plant can't make artillery shells, but we're not shutting it down...
Again, totally non-researched, off the top of my head assumption.
Where does the Thorium come from and how abundant is it?
http://www.forbes.com/sites/energysource/2012/02/16/the-thin...
[Edit] corrected use of "fissile" -> "fertile"
While strictly correct, this is extremely misleading. Current resources would be exhausted in a decade, but resources are defined as the known deposits extractable under current market prices. Should we actually start using a lot of uranium, the price would spike, which would lead to a lot more deposits becoming economically available. This has almost no effect on the cost of nuclear power, as the cost of the raw uranium isn't a large part of the cost of producing power.
The end game there is when the price rises sufficiently for extraction from seawater becoming profitable. The world's seas have ~1000 times more uranium than conventional ground-based resources.
Nuclear fuels will not run out in this millennium.
For a real comparison, you should look at fast reactors, which use the rest of the uranium. That takes your estimate up to about a thousand years. But the estimate is looking at economically recoverable reserves, and if the same ore produces a hundred times as much energy, a lot more becomes economically recoverable.
Thorium would be 3-4 times more abundant than that. (But if seawater extraction works out, uranium will have the advantage again.)
If there ever appears a new intelligent civilization millions of years after humanity, they'll be very disappointed to be on an Earth without any Uranium or Thorium!
Uranium will simply not run out. The "10 years" figure stems entirely from misunderstanding what the word "reserve" means in mining.
Also, there is 3-4 times more natural thorium than natural uranium. Of natural uranium only 0.7% is U-235, which is useful in a traditional reactor. Breeders can use all of it. Thorium reactors are all breeders. Assuming only resources extractable at current market prices, using uranium breeders the reserves would last ~1400 years and the thorium reserves would last ~5k years.
